doi: 10.18282/amor.v3.is1.211
ORIGINAL RESEARCH ARTICLE
Healthcare-associated infections among patients with different types of
acute leukemia in China: A surveillance-based study
Yunhong Liu1,2, Tianyuan Yan3, Jingna Wang4, Carmen WH Chan1*, Doris YP Leung1, Shuhui
Wang2*
The Nethersole School of Nursing, The Chinese University of Hong Kong, Hong Kong SAR, China
The Department of Infection Prevention and Control, Qilu Hospital of Shandong University, Jinan, Shandong Province, China
3
The Nursing School of Shandong University, Jinan, Shandong Province, China
4
Neonatal Intensive Care Unit, Qilu Hospital of Shandong University, Jinan, Shandong Province, China
1
2
Abstract: Healthcare-associated infections (HAIs) may be diverse among patients with acute myeloid leukemia (AML) and
acute lymphoblastic leukemia (ALL). However studies were limited. We aimed to quantify and compare the characteristics
and relevant factors associated with HAIs among patients with AML and ALL. The surveillance-based study was conducted
in a university teaching hospital. All adult patients diagnosed with acute leukemia (AL) and admitted into the Hematology
Department for over 48 h were included. HAI characteristics and relevant factors were compared between AML and ALL
patients. A consecutive sample of 994 patients with AL was recruited. The proportions of infected cases (27.78% versus
28.31%, p = 0.888) and the HAI incidence (21.29 versus 22.82 per 1,000 patient-days at risk, p = 0.733) were comparable
among AML and ALL patients, respectively. Compared with ALL patients, higher risks of HAIs were found among AML
patients with increasing duration of chemotherapy or lower hemoglobin level. Meanwhile, increased length of stays during
previous cycles of chemotherapy, lower level of platelets, and diabetes were associated with higher risks of HAIs among
ALL patients compared with AML patients. In conclusion, our results found that AML and ALL patients experienced
different risks of HAIs associated with diverse relevant factors. Future multi-center studies are needed to provide stronger
evidence.
Keywords: acute myeloid leukemia; acute lymphoblastic leukemia; healthcare-associated infections; relevant factors of
infections
Citation: Liu Y, Yan T, Wang J, Chan CWH, Leung DYP, et al. Healthcare-associated infections among patients with
different types of acute leukemia in China: A surveillance-based study. Adv Mod Oncol Res 2017; 3(S1): 79–88.
http://dx.doi.org/10.18282/amor.v3.is1.211.
*Correspondence to:
Shuhui Wang, The Department of Infection Prevention and Control, Qilu Hospital of Shandong University, Jinan, Shandong
Province, China; [email protected]
Carmen W.H. Chan, The Nethersole School of Nursing, The Chinese University of Hong Kong, Hong Kong SAR, China;
[email protected].
Received: 24th March 2017; Accepted: 17th April 2017; Published Online: 28th April 2017
Introduction
Healthcare-associated infections (HAIs) affect many
patients worldwide. They are closely associated with longer
hospital stays, increased hospitalization costs, and even
higher mortality[1]. According to the European Centers for
Disease Prevention and Control (ECDC), approximately
4.1 million patients develop HAI in Europe each year[2]. In
China, HAIs have also been shown to be prevalent. One
study in China found that the incidence of HAIs was as high
as 30.78% among adult patients with acute leukemia (AL)[3].
Indeed, the huge populations in major Chinese cities have
prompted the establishment of a number of super hospitals
Copyright © 2017 Liu Y, et. al. This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0
International License (http://creativecommons.org/licenses/by-nc/4.0/), permitting all non-commercial use, distribution, and reproduction in any medium,
provided the original work is properly cited.
79
Healthcare-associated infections among patients with different types of acute leukemia in China: A surveillance-based study
that are able to accommodate more than 4,500 patients.
These super hospitals tend to be overcrowded, with their
medical staff working under immense pressure. These
would impose challenges to effective HAI surveillance
and control in these hospitals. In order to overcome these
challenges, a better understanding on the causal factors of
HAIs is required.
AL, a type of severe hematological malignancy, can
be classified into two different categories—acute myeloid
leukemia (AML) and acute lymphoblastic leukemia
(ALL)—based on the lineage of the malignant cells[4]. The
French-American-British (FAB) classification system has
divided AML into eight subtypes (M0 through M7), and
ALL into three subtypes (L1 through L3), based on the
morphology and degree of maturity of the cancer cells[5].
FAB classification has been widely used around the world
eventhough a new classification system was developed
by the World Health Organization (WHO) [6]. The rapid
development of medical technology has contributed to
a significant progress in the development of effective
therapeutic treatment of AL, which greatly reduces the
overall mortality among AL patients[7,8]. Nevertheless, these
patients are highly vulnerable to the development of HAIs
due to underlying diseases such as diabetes, prolonged
neutropenia (absolute neutrophil count <1.5×109 cells/L),
aggressive treatment strategies and medication use[9,10], and
more importantly, the use of chemotherapy[11,12]. Intense
chemotherapy using cytotoxic drugs may suppress the
immune system, leading to myelosuppression and an
increased risk of HAI development[13].
Most previous studies have tested the etiological
pathogens, such as types of bacteria and fungus of HAIs,
and identified the potential risk factors for certain groups
of AL patients such as bone marrow transplant patients
with AL, pediatric AL patients, AML patients or ALL
patients[14-20]. Risk factors that they have identified include
neutropenia, chemotherapy, catheterization and so on.
Furthermore, a few studies revealed that AML patients
are more prone to the development of pneumonia than
ALL patients. Huoi et al.[9] found that hospital-acquired
pneumonia incidence was 2.4 per 1,000 patient-days higher
among AML patients compared to ALL patients. Garcia
et al.[21] also found that the 28-day cumulative incidence
rate of pneumonia was 14.5% higher among patients with
AML than those with ALL. Nevertheless, little are known
about the differences in the characteristics and the relevant
factors associated with HAIs experienced by AML and
80
ALL patients. Identifying individuals with high HAI risks
and discovering the relevant factors associated with HAIs
would help facilitate the development of preventative
strategies, so as to minimize the occurrence of HAIs among
AL patients.
The aim of this study is to compare the different HAI
characteristics and relevant factors among AML and
ALL patients. The primary objective of our study is to
make comparisons on the proportions of infected cases,
HAI incidence and relevant factors associated with HAIs
between AML and ALL patients. The secondary objective
is to investigate and compare the common infection types
among AML and ALL patients. We hypothesize that
patients with AML would be at higher risks of HAIs than
those with ALL.
Materials and methods
Design and participants
This study involved secondary data analysis. The original
study, aiming at examining relevant factors of HAIs among
AL patients, was a hospital surveillance-based study
through administrative-monitored electronic databases.
In three consecutive years, all adult patients (identified
via administrative-monitored electronic databases and
who were admitted for over 48 h into the Hematology
Department of a comprehensive tertiary hospital affiliated
with a medical school) were included in the study.
Those with severe major organ dysfunction and serious
complications were excluded. AL patients who previously
developed chronic leukemia or myelodysplastic syndrome
were also excluded. Categories of AL (AML or ALL)
were confirmed by cytogenetic tests, analyses of bone
marrow morphology, molecular detection of oncogenic
protein, and clinical manifestations. For patients who
cannot be classified into a particular subtype of ALL during
hospitalization, we classified them as uncertainty.
In this study, all 994 participants of the original study
were included in the analysis. Using the commonly
accepted rule of thumb of sample size for logistic regression
(20 subjects per variable), the sample size was sufficient to
ensure an acceptable statistical power for logistic regression
models with at most 50 variables.
Definition of HAIs
HAIs were recognized primarily through their definition
doi:10.18282/amor.v3.is1.211
Liu Y, et. al.
published by the Centers for Disease Control and
Prevention[22]. Diagnosis of HAIs depends on two main
factors: First, the necessary combination of clinical
manifestations (such as fever and cough), laboratory results
(such as blood test and bacteria culture) and other diagnostic
tests (such as X-ray) should be considered. Second, only
when HAIs occurred at least 48 h after hospital admission
would the patients be considered to have hospital-acquired
HAIs. Types of HAIs included in this particular study were
upper and lower respiratory tract infection, oral infection,
skin or soft tissue infection, urinary tract infection (UTI),
gastrointestinal tract infection (GTI), bloodstream infection
(BSI) and other infections, including all infection types
apart from the mentioned above.
Data collection
Data were extracted from a prospectively designed database
named Hospital Infection Surveillance System and Hospital
Information System by two trained staff members in
infection control, using a self-made information extraction
form. Data on five different aspects were collected: (1)
demographic characteristics; (2) clinical information, i.e.,
length of stay (LOS), admission and discharge diagnosis,
underlying disease and the use of peripherally inserted
central catheter (PICC); (3) AL characteristics (types and
subtypes of AL, central nervous system leukemia and
marrow proliferation analysis); (4) laboratory test results
(routine blood tests) and; (5) details of HAIs (infection
on admission, the day of infection and infection types).
Finally, 16 potential HAI factors (10 were continuous and
6 were categorical variables) were included in the logistic
regression analysis, including body mass index (BMI),
LOS, infection on admission, blood transfusion, underlying
diseases (for e.g., diabetes), chemotherapy (utilization
of chemotherapy, duration of chemotherapy, previous
cycles of chemotherapy, cycles of chemotherapy during
hospital stay and stage of chemotherapy), blood tests
(amount of platelets, white blood cells, hemoglobin and
neutrophil granulocyte in blood), and marrow proliferation
analysis[21,23-26].
Data analysis
Data were analyzed by SPSS version 22.0 (SPSS, Inc.,
Chicago, Illinois, USA). Mean ± standard deviation
(SD), median and quartile interval (QI), or frequencies
and constitute proportions were used for summarizing the
doi:10.18282/amor.v3.is1.211
characteristics of the sample. The proportions of infected
cases were expressed as a percentage of the number of
patients with HAIs out of the total number of patients
within a particular AL types, and the HAI incidence was
expressed as the number of HAIs episodes per 1,000
patient-days of a particular AL types. Then, we created
a new variable ‘AML’ to indicate the two types of AL
with the score of 1 representing AML while the score of 0
representing ALL.
Chi-square test, Mann-Whitney U test or T-test was used
for comparison of the variables between the two AL types.
In order to identify the independent relevant factors for
HAIs by AL types, we included the interaction terms of the
16 potential factors with AML. Accordingly, all the main
effects and interaction terms of the 16 potential factors were
included in logistic regression analyses using backward
stepwise (likelihood ratio) method. Variables were included
with p < 0.05 and excluded with p > 0.1. For the final
logistic regression model, results were presented as odds
ratio (OR) and 95% confidence intervals (95% CI), and p <
0.05 was considered statistically significant.
Ethics statement
The research protocol was approved by the Ethics
Committee of the selected hospital. The collected data
remain strictly confidential.
Results
The characteristics and general information of
AML and ALL patients
Throughout the study, 994 AL patients were determined
to meet the eligibility criteria, and were enrolled into the
study among the 1,573 patients admitted to the Department
of Hematology. The study participants ranged in age
from 18 to 84 years old (42.25 ± 16.01) and 565 (56.8%)
participants were male. Among the enrolled individuals,
828 (83.30%) were AML patients, and 166 (16.70%) were
ALL patients. A flowchart depicting the recruitment process
of the study is presented in Figure 1.
Generally, ALL patients were significantly younger
and had a lower BMI. As shown in Table 1, there were
proportionately more male patients in the ALL group
compared with the AML group (65.7% versus 55.1%,
respectively, p = 0.007). In addition, more ALL patients
were in the induction treatment stage of chemotherapy. A
81
Healthcare-associated infections among patients with different types of acute leukemia in China: A surveillance-based study
Figure 1. A CONSORT diagram showing the flow of participant recruitment for the study
larger proportion of ALL patients used stimulating factors,
hormone, immuno-modulators and peripherally inserted
central catheter (PICC) for cancer treatment. More ALL
patients were diagnosed with polar hyperplasia and fewer
of them were diagnosed with active proliferation in the
marrow proliferation analysis. All the above differences
were statistically significant between patients with AML
and ALL (p < 0.05). Other categories were comparable
between the two groups.
HAIs characteristics in different types of AL
patients
Table 2 presents the characteristics of HAIs acquired by
different types and subtypes of AL patients in our study.
Among the 994 patients, 277 had acquired HAIs, with a
total of 306 HAI episodes. The proportion of infected cases
was 27.9%, and that of HAI incidence reached 21.52% (95%
CI: 21.24%–21.80%). The differences in the proportions of
both infected cases (27.78% versus 28.31%, p = 0.888) and
HAI incidence (21.29% versus 22.82% patient-days, p =
0.733) were not statistically significant between AML and
ALL groups, respectively.
Table 3 shows that the most common types of infection
among AML and ALL patients were upper respiratory
tract infection (as a proportion of all infection types,
43.39% in AML group and 46.81% in ALL group, p >
0.05) and pneumonia (36.36% in AML group and 31.91%
in ALL group, p > 0.05), followed by oral infection and
82
gastrointestinal tract infection. The types of infection in this
study were not significantly different between AML and
ALL patients.
Relevant factors associated with the
development of HAIs among patients with
AML and ALL
Through adjusted logistic regression analysis, Table
4 shows that five factors including LOS, infection on
admission, utilization of chemotherapy, duration of
chemotherapy and hemoglobin concentration were factors
associated with HAIs in both AML and ALL groups.
Specifically, the utilization of chemotherapy and prolonged
LOS were associated with increased risks of HAIs (OR
>1, p < 0.01), while infection on admission, prolonged
duration of chemotherapy and higher level of hemoglobin
were associated with lower risk of HAIs (OR <1, p <
0.01). For the patients with diabetes, the risks of HAIs
were significantly lower among AML patients compared
with ALL patients (OR = 0.279; 95% CI: 0.113~0.691;
p = 0.006). With regard to other variables, i.e., increased
hospitalization day (OR = 0.797; 95% CI: 0.714~0.890; p
< 0.001), more chemotherapy cycle undertaken previously
(OR = 0.939; 95% CI: 0.895~0.985; p = 0.010) and higher
level of platelets (OR = 0.995; 95% CI: 0.990~0.999; p
= 0.025), the risks of HAIs were also lower among AML
patients. On the other hand, compared with ALL patients,
the risks of HAIs were significantly higher among AML
doi:10.18282/amor.v3.is1.211
Liu Y, et. al.
Table 1. Demographics and clinical characteristics of acute myeloid leukemia and acute lymphoid leukemia patients
Variables
AML (N = 828)
ALL (N = 166)
p
43 ± 15.68
33.5 ± 16.84
<0.001**
Male
456 (55.1%)
109 (65.7%)
0.007**
BMI (kg/m2)
24.84 ± 3.34
23.55 ± 3.35
<0.001**
Ages (years)
Central nervous system leukemia
17 (2.1%)
3 (1.8%)
0.837
Infection on admission
180 (21.7%)
43 (25.9%)
0.240
HAIs
230 (27.8%)
47 (28.3%)
0.888
Blood transfusion
390 (47.1%)
81 (48.8%)
0.690
LOS (days)
11 (7–20)
10 (7–17)
0.118
Death
17 (2.1%)
3 (1.8%)
0.837
Chemotherapy usage
693 (83.7%)
132 (79.5%)
0.191
Duration of chemotherapy
5.80 ± 4.05
5.10 ± 4.24
0.045*
Previous cycles of chemotherapy
4.44 ± 4.18
3.14 ± 3.17
<0.001**
Cycles of chemotherapy during this hospital stay
0.86 ± 0.40
0.80 ± 0.40
0.097
Chemotherapy:
Stage of chemotherapy:
Induction treatment
182 (22.0%)
58 (34.9%)
Consolidation
502 (60.6%)
89 (53.6%)
Relapsed or refractory period
144 (17.4%)
19 (11.4%)
Hypertension
60 (7.2%)
18 (10.8%)
0.116
Diabetes
72 (8.7%)
11 (6.6%)
0.379
Cardiovascular disease
15 (1.8%)
4 (2.4%)
0.608
Hyperlipemia
5 (0.6%)
0
0.316
Tuberculosis
2 (0.2%)
0
0.526
0.57 ± 0.032
0.48 ± 0.074
0.239
331 (40.0%)
86 (51.8%)
0.005**
41 (5.0%)
18 (10.8%)
0.003**
Immuno-modulator
508 (61.4%)
118 (71.1%)
0.018*
Antibiotics
415 (50.1%)
92 (55.4%)
0.212
98 (11.8%)
30 (18.1%)
0.029*
4 (0.5%)
1 (0.6%)
0.843
0.001**
Underlying disease:
Number of underlying diseases
Medication usage:
Stimulating factor
Hormone
Catheterization:
PICC
Urinary catheterization
Marrow Proliferation Analysis:
Polar hyperplasia
61 (7.4%)
27 (16.3%)
Obvious proliferation
99 (12.0%)
23 (13.9%)
Active proliferation
552 (66.7%)
87 (52.4%)
Reduced or extremely reduced proliferation
20 (2.4%)
6 (3.6%)
Without marrow analysis
96 (11.6%)
23 (13.9%)
3.16 (2.39–3.82)
3.30 (2.59–3.94)
0.001**
Routine blood test:
Red blood cell (×1012 cells/L)
9
White blood cell (×10 cells/L)
Platelets (×109 cells/L)
Hemoglobin (g/L)
Neutrophil granulocyte (×109 cells/L)
0.129
3.48 (1.75–5.81)
3.26 (1.07–5.68)
0.271
115 (26–218)
140 (30.75–265.5)
0.073
96 (71–118)
96 (75.75–114)
0.919
1.84 (0.56–3.43)
1.78 (0.43–3.60)
0.735
Note: *p < 0.05; **p < 0.01. Values are presented either as mean ± standard deviation, N (%), or median (interquartile range). Abbreviations: BMI: body
mass index; HAIs: hospital acquired infections; LOS: length of stay; PICC: peripherally inserted central catheter
doi:10.18282/amor.v3.is1.211
83
Healthcare-associated infections among patients with different types of acute leukemia in China: A surveillance-based study
Table 2. The association between different types of acute leukemia and hospital-acquired infections
Types and subtypes of AL
AML
N
The proportions of infected
cases
N
%
Cumulative
(average)
patient-days
HAI incidence
Episodes
% patient-days
(95% CI)
21.29 (20.97–21.61)
828
230
27.78
12,073 (15)
257
M0
9
3
33.33
101 (10)
3
29.70
M1
17
8
47.06
342 (20)
8
23.39 (10.14–36.64)
M2
75
16
21.33
1,003 (13)
18
17.95 (14.48–21.41)
M3
240
46
19.17
2,912 (12)
48
16.48 (15.50–17.46)
22.98 (20.95–25.02)
M4
140
41
29.29
2,045 (15)
47
M5
325
105
32.31
5,322 (16)
121
22.74 (21.88–23.60)
M6
22
11
50.00
348 (16)
12
34.48 (21.14–47.83)
M7
ALL
0
0
0
0
0
0
166
47
28.31
2,147 (13)
49
22.82 (21.22–24.43)
L1
10
3
30.00
185 (19)
3
16.22
L2
55
23
41.82
777 (14)
25
32.18 (26.76–37.59)
L3
7
2
28.57
101 (14)
2
19.80
94
19
20.21
1084(12)
19
17.53 (15.03–20.03)
994
277
27.90
14220(14)
306
21.52 (21.24–21.80)
Uncertainty
Total
Table 3. Comparisons of hospital-acquired infection types in infected acute myeloid leukemia and acute lymphoid leukemia patients
Total
(N, %)
AML
(N, %)
Upper respiratory tract infection
127 (43.94)
Pneumonia/Lower respiratory tract infection
103 (35.64)
Infection types
ALL
(N, %)
p
105 (43.39)
22 (46.81)
0.840
88 (36.36)
15 (31.91)
0.539
Oral infection
17 (5.88)
15 (6.20)
2 (4.26)
0.441
Skin/Soft tissue infection
14 (4.84)
13 (5.37)
1 (2.13)
0.293
UTI
3 (1.04)
2 (0.83)
1 (2.13)
0.422
GTI
16 (5.54)
13 (5.37)
3 (6.38)
0.516
BSI
9 (3.11)
6 (2.48)
3 (6.38)
0.178
Other infections
11 (3.81)
9 (3.72)
2 (4.26)
0.572
289 (100.00)
242 (100.00)
47 (100.00)
0.888
Total
UTI: Urinary tract infection; GTI: gastrointestinal tract infection; BSI: bloodstream infection
Table 4. Logistic regression of adjusted main relevant factors and interaction terms by AL types for the development of hospital-acquired
infections among AL patients
Variables
OR (95% CI)
p
LOS (days)
1.382 (1.238~1.544)
<0.001**
Infection on admission (Yes/No)
0.142 (0.084~0.241)
<0.001**
Chemotherapy (Yes/No)
6.238 (2.807~13.864)
<0.001**
Duration of chemotherapy
0.709 (0.601~0.837)
<0.001**
Main effects:
Platelets
0.997 (0.993~1.001)
0.122
Hemoglobin
0.970 (0.954~0.985)
<0.001**
LOS by AML
0.797 (0.714~0.890)
<0.001**
Diabetes (Yes/No) by AML
0.279 (0.113~0.691)
0.006**
Duration of chemotherapy by AML
1.329 (1.120~1.577)
0.001**
Previous cycles of chemotherapy by AML
0.939 (0.895~0.985)
0.010*
Platelets by AML
0.995 (0.990~0.999)
0.025*
Hemoglobin by AML
1.024 (1.008~1.040)
0.003**
Interaction terms:
**
*
Note: p < 0.01; p < 0.05. Abbreviations: LOS: length of stay. The AL types in this analysis were coded as 1 = AML, and 0 = ALL. The hospital-acquired
infection (HAI) was coded as 1 = with HAI, and 0 = without HAI.
84
doi:10.18282/amor.v3.is1.211
Liu Y, et. al.
patients with prolonged duration of chemotherapy (OR =
1.329; 95% CI: 1.120~1.577; p = 0.001) and lower level of
hemoglobin (OR = 1.024; 95% CI: 1.008~1.040; p = 0.003).
Discussion
The primary objective of this retrospective study is
to compare the proportions of infected cases and HAI
incidence, and the relevant factors associated with HAIs
between AML and ALL patient groups. Our findings
suggest that the proportion of infected cases and HAI
incidence were comparable among AML and ALL patients.
In our study, the overall proportion of infected cases
was found to be 27.90% and the total HAI incidence was
21.52% patient-days, which were higher than the study
on hematological malignancies[23], hence this indicates the
severe condition of HAIs among AL patients. Interestingly,
our data appear to show somewhat contradictions to our
hypothesis that AML patients would be at higher risk to
develop HAIs than ALL patients. This could be explained
by a number of factors, including the use of medication
and PICC, bone marrow hyperplasia and induction
chemotherapy. First, ALL patients in our study took
medication more frequently than AML patients, including
stimulating factors, hormone and immuno-modulators. This
might leave them at higher risk of developing HAIs, as
previous studies showed that the use of steroid hormones
medication is likely to result in HAIs in ALL patients[27].
Second, more patients with ALL were inserted with a
PICC, which was previously shown to lead to patients’
higher susceptibility to bloodstream infection (BSI)[28]. In
our study, the proportion of BSI was indeed higher in ALL
group than AML group (6.38% versus 2.48%, respectively)
though the difference was not significant. Third, more
ALL patients were in a condition of polar hyperplasia,
in which the bone marrow cells produce a large number
of immature blood cells. These conditions suggest poor
health status and a weakened immune system, which
make patients more susceptible to HAIs. Fourth, more
ALL patients undertaking chemotherapy in this study were
under the induction chemotherapy stage. Previous studies
have demonstrated that neutropenia usually occurs during
the first course of induction chemotherapy[24], and this
increases the risks of infections among patients undertaking
induction chemotherapy [24,28,29] . These factors could
contribute to slightly higher proportions of infected cases
and HAI incidence among ALL patients observed in this
doi:10.18282/amor.v3.is1.211
study. As far as HAI types are concerned, upper respiratory
tract infection and pneumonia were the top two common
types of infection in our study, which was consistent
with previous findings. Garcia et al.[21] demonstrated that
pneumonia continues to be a major HAI problem, which
was associated with significant morbidity, mortality, and
health-care resource utilization among AL patients.
As for relevant factors of HAIs, various relevant factors
and interaction terms are found to be associated with HAIs.
Prolonged LOS is a risk factor for HAIs, and compared
with AML patients, patients with ALL are at higher risks.
Recently, Ford et al. [1] found that LOS was associated
with the gastrointestinal colonization of vancomycinresistant Enterococcus (VRE). Therefore, AL patients
are more prone to infections by various pathogens in a
hospital environment, owing to their impaired immune
function. Our finding that infection on admission could
decrease the risks of HAIs is in contrast to a previous
study[23]. The reason may lie in the fact that patients with
infections on admission were given antibiotic treatment
during hospitalization and this could protect them from
further HAIs. Furthermore, it is well established that
chemotherapy is very likely to weaken the immune system
and is associated with more HAIs. As chemotherapeutic
drugs can cause damage to the bone marrow, they lead to
the interference to the production of sufficient red blood
cells, white blood cells, and platelets. However, in terms
of prolonged duration of chemotherapy, AML patients
have significantly greater risks of HAIs than ALL patients.
The reason may be that AML treatment usually causes
a more prolonged period of neutropenia and AML itself
is a risk factor for pneumonia, even after adjusted for
neutropenia[21]. On the other hand, one study showed that
the first course of induction chemotherapy is the stage when
HAIs are most likely to occur[24]. Subsequently, as patients
progress through the chemotherapy regimen, their risks
of developing HAIs decrease. A recent study suggested
that platelets play a vital role in inflammation and immune
response [25]. Considering that severe thrombocytopenia
signifies severity of leukemia, a higher platelets level could
be linked to decreased incidence of HAIs[21]. In addition, AL
patients may also suffer from hemoglobin deficiency, which
may be caused by bone marrow failure and side effects of
chemotherapy. This condition leads to the development of
symptoms associated with anemia (hemoglobin <110 g/L),
which would weaken the immune system[30]. Furthermore,
the higher HAI risks associated with diabetes could be due
85
Healthcare-associated infections among patients with different types of acute leukemia in China: A surveillance-based study
to high blood sugar levels and immunosuppression. Indeed,
Dare et al.[31] found that hyperglycemia is associated with
higher risks of bacterial or fungal infections, thereby
providing a link between diabetes, raised blood sugar
levels and HAI risks. There is also an evidence of a dosedependent response between mortality and increasing
hyperglycemia among AML patients[32].
There are a few limitations in our study. First, we
used logistic regression to identify the relevant factors
associated with HAIs, but this analytical method did not
take into account of the possibility of changes in these
relevant factors over time. This potential drawback could
be resolved by using other statistical methods, such as
proportional hazards model which considers the risks
of event change over time. Second, only sixteen of the
factors previously demonstrated to be the potential factors
of HAIs were used in our analysis, and they may not
necessarily represent all the factors that are associated with
HAIs. Third, the data were retrospectively collected from
hospital information databases and did not include postdischarge follow-up information. Fourth, this study was a
Author contributions
The study design was done by J Wang and S Wang. Data
collection and analysis were done J Wang and DYP Leung.
Y Liu and T Yan wrote the manuscript while CWH Chan
supervised the data analysis and result presentation. The
manuscript was critically revised by CWH Chan. Lastly,
statistical consultation was done by DYP Leung.
Acknowledgements
We would like to thank the staff members from Department
of Hematology and Medical Records Room of Qilu
Hospital affiliated with Shandong University, Shandong
Province, China, for providing assistance with data
acquisition. We would also like to thank Mr. Bernard MH
Law for his help in language editing.
Conflict of interest
The authors declare no potential conflict of interest with
respect to the research, authorship, and/or publication of
this article.
single-center study, with all of the recruited patients from a
References
university teaching hospital, thus the generalizability of our
1. Ford CD, Lopansri BK, Haydoura S, Snow G, Dascomb KK,
findings to other medical settings cannot be assumed.
Conclusion
In conclusion, we demonstrated that the situation of HAIs
was still severe among AL patients. The proportions of
infected cases and the HAI incidence were comparable
among AML and ALL patients. Duration of chemotherapy
and amount of hemoglobin were associated with higher
risks of HAIs among AML patients compared with ALL
patients. Meanwhile, length of stays, previous cycles
of chemotherapy, diabetes and amount of platelets
were associated with higher risks of HAIs among ALL
patients compared with AML patients. Further multicenter prospective studies are required to provide stronger
evidence for the association between HAIs and AL types.
Future studies on the susceptibility of AL patients to HAIs
may be performed using a multi-disciplinary approach,
et al. Frequency, risk factors, and outcomes of vancomycinresistant enterococcus colonization and infection in patients
with newly diagnosed acute leukemia: Different patterns in
patients with acute myelogenous and acute lymphoblastic
leukemia. Infect Control Hosp Epidemiol 2015; 36(1): 47−53.
doi: 10.1017/ice.2014.3.
2. Healthcare-associated infections [Internet]. Sweden: European
Centre for Disease Prevention and Control; 2005–2017 [cited
2017 March 23]. Available from: http:// http://ecdc.europa.eu/
en/healthtopics/Healthcare-associated_infections/Pages/index.
aspx.
3. Wang J. Study on the risk factors and direct economic
losses of nosocomial infection with acute leukemia patients
[Dissertation]. China: Shandong University; 2014.
4. Sabattini E, Bacci F, Sagramoso C, Pileri SA. WHO
classification of tumours of haematopoietic and lymphoid
tissues in 2008: An overview. Pathologica 2010; 102(3):
83−87.
including the conduction of infection gene detection,
5. Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton
molecular biology research, or an investigation on the
DAG, et al. Proposals for the classification of the acute
possibility of the use of alternative medicine to treat
leukaemias. French-American-British (FAB) co-operative
infections.
group. Br J Haematol 1976; 33(4): 451−458. doi: 10.1111/
86
doi:10.18282/amor.v3.is1.211
Liu Y, et. al.
j.1365-2141.1976.tb03563.x.
16.Dettenkofer M, Wenzler-Röttele S, Babikir R, Bertz H, Ebner
6. DeAngelo DJ, Pui C. Acute lymphoblastic leukemia and
W, et al. Surveillance of nosocomial sepsis and pneumonia
lymphoblastic lymphoma. 6th ed. In: American Society of
in patients with a bone marrow or peripheral blood stem cell
Hematology Self-Assessment Program. Washington: American
transplant: A multicenter project. Clin Infect Dis 2005; 40(9):
Society of Hematology. 2013. p. 491–507.
926−931. doi: 10.1086/428046.
7. Gooley TA, Chien JW, Pergam SA, Hingorani S, Sorror ML,
17.Al-Tonbary YA, Soliman OE, Sarhan MM, Hegazi MA, El-
et al. Reduced mortality after allogeneic hematopoietic-cell
Ashry RA, et al. Nosocomial infections and fever of unknown
transplantation. N Engl J Med 2010; 363(22): 2091−2101. doi:
origin in pediatric hematology/oncology unit: A retrospective
10.1056/NEJMoa1004383.
annual study. World J Pediatr 2011; 7(1): 60−64. doi: 10.1007/
8. Pulte D, Gondos A, Brenner H. Trends in 5- and 10-
s12519-010-0212-1.
year survival after diagnosis with childhood hematologic
18.Semochkin SV, Tolstykh TN, Arkhipova NV, Ivanova VL,
malignancies in the United States, 1990–2004. J Natl Cancer
Kliueva OV, et al. Clinical and epidemiological characteristics
Inst 2008; 100(18): 1301−1309. doi: 10.1093/jnci/djn276.
of acute myeloid leukemias in adults according to the data
9. Huoi C, Vanhems P, Nicolle MC, Michallet M, Benet T.
of municipal hematology departments in Moscow. Ter Arkh
Incidence of hospital-acquired pneumonia, bacteraemia
2015; 87(7): 26−32. doi: 10.17116/terarkh201587726-32.
and urinary tract infections in patients with haematological
19.Giri S, Chi M, Johnson B, McCormick D, Jamy O, et al.
malignancies, 2004-2010: A surveillance-based study. PLoS
Secondary acute lymphoblastic leukemia is an independent
One 2013; 8(3): e58121. doi: 10.1371/journal.pone.0058121.
predictor of poor prognosis. Leuk Res 2015; 39(12):
10.Thirumala R, Ramaswamy M, Chawla S. Diagnosis and
1342−1346. doi: 10.1016/j.leukres.2015.09.011.
management of infectious complications in critically ill
20.Karachunskiĭ AI, Rumiantseva IuV, Lagoĭko SN, Bührer C,
patients with cancer. Crit Care Clin 2010; 26(1): 59−91. doi:
Tallen G, et al. Age-related characteristics of the efficacy
10.1016/j.ccc.2009.09.007.
of different glucocorti-costeroids in the therapy of acute
11. Crawford J, Dale DC, Lyman GH. Chemotherapy-induced
lymphoblastic leukemia. Ter Arkh 2015; 87(7): 41−50.
neutropenia: Risks, consequences, and new directions for its
21.Garcia JB, Lei X, Wierda W, Cortes JE, Dickey BF, et al.
management. Cancer 2004; 100(2): 228−237. doi: 10.1002/
Pneumonia during remission induction chemotherapy in
cncr.11882.
patients with acute leukemia. Ann Am Thorac Soc 2013;
12.Alatorre-Fernández P, Mayoral-Terán C, Velázquez-Acosta
10(5): 432−440. doi: 10.1513/AnnalsATS.201304-097OC.
C, Franco-Rodríguez C, Flores-Moreno K, et al. A polyclonal
22.Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance
outbreak of bloodstream infections by Enterococcus faecium in
definition of health care-associated infection and criteria
patients with hematologic malignancies. Am J Infect Control
for specific types of infections in the acute care setting.
2017; 45(3): 260−266. doi: 10.1016/j.ajic.2016.10.002.
Am J Infect Control 2008; 36(5): 309−332. doi: 10.1016/
13.Bow EJ, Rotstein C, Noskin GA, Laverdiere M, Schwarer
j.ajic.2008.03.002.
AP, et al. A randomized, open-label, multicenter comparative
23.Liu H, Zhao J, Xing Y, Li M, Du M, et al. Nosocomial
study of the efficacy and safety of piperacillin-tazobactam and
infection in adult admissions with hematological malignancies
cefepime for the empirical treatment of febrile neutropenic
originating from different lineages: A prospective
episodes in patients with hematologic malignancies. Clin
observational study. PLoS One 2014; 9(11): e113506. doi:
Infect Dis 2006; 43(4): 447−459. doi: 10.1086/505393.
10.1371/journal.pone.0113506.
14.Dettenkofer M, Ebner W, Bertz H, Babikir R, Finke U, et al.
24.Biswal S, Godnaik C. Incidence and management of infections
Surveillance of nosocomial infections in adult recipients of
in patients with acute leukemia following chemotherapy in
allogeneic and autologous bone marrow and peripheral blood
general wards. Ecancermedicalscience 2013; 7: 310. doi:
stem-cell transplantation. Bone Marrow Transpl 2003; 31(9):
10.3332/ecancer.2013.310.
795−801. doi: 10.1038/sj.bmt.1703920.
25.Li C, Li J, Li Y, Lang S, Yougbare I, et al. Crosstalk
15.Urrea M, Rives S, Cruz O, Navarro A, Garcia JJ, et al.
between platelets and the immune system: Old systems
Nosocomial infections among pediatric hematology/
with new discoveries. Adv Hematol 2012: 384685.doi:
oncology patients: Results of a prospective incidence study.
10.1155/2012/384685.
Am J Infect Control 2004; 32(4): 205−208. doi: 10.1016/
j.ajic.2003.10.013.
doi:10.18282/amor.v3.is1.211
26.Ilavská S, Horváthová M, Szabová M, Nemessányi T, Jahnová
E, et al. Association between the human immune response and
87
Healthcare-associated infections among patients with different types of acute leukemia in China: A surveillance-based study
body mass index. Hum Immunol 2012; 73(5): 480−485. doi:
hematologic malignancies. Leukemia 2005; 19(4): 545−550.
10.1016/j.humimm.2012.02.023.
doi: 10.1038/sj.leu.2403674.
27.Shafey A, Ethier MC, Traubici J, Naqvi A, Sung L. Incidence,
30.Ekiz C, Agaoglu L, Karakas Z, Gurel N, Yalcin I. The effect of
risk factors, and outcomes of enteritis, typhlitis, and colitis
iron deficiency anemia on the function of the immune system.
in children with acute leukemia. Int J Pediatr Hematol Oncol
Hematol J 2005; 5(7): 579−583. doi: 10.1038/sj.thj.6200574.
2013; 35(7): 514−517. doi: 10.1097/MPH.0b013e31829f3259.
31.Dare JM, Moppett JP, Shield JP, Hunt LP, Stevens MC. The
28.Berrueco R, Rives S, Catala A, Toll T, Gene A, et al. Pro­
impact of hyperglycemia on risk of infection and early death
spective surveillance study of blood stream infections
during induction therapy for acute lymphoblastic leukemia
associated with central venous access devices (port-type) in
(ALL). Pediatr Blood Cancer 2013; 60(12): e157−e159. doi:
children with acute leukemia: An intervention program. J
Pediatr Hematol Oncol 2013; 35(5): e194–e199. doi: 10.1097/
MPH.0b013e318290c24f.
29.Mühlemann K, Wenger C, Zenhäusern R, Täuber MG. Risk
factors for invasive aspergillosis in neutropenic patients with
88
10.1002/pbc.24689.
32.Ali NA, O’Brien JM, Blum W, Byrd JC, Klisovic RB, et al.
Hyperglycemia in patients with acute myeloid leukemia is
associated with increased hospital mortality. Cancer 2007;
110(1): 96−102. doi: 10.1002/cncr.22777.
doi:10.18282/amor.v3.is1.211